Effects of Surgery and Radiotherapy on Recurrent Skull Base Meningiomas: Clinical and Biological Analyses

Abstract

We reviewed the medical records of 392 patients who underwent initial surgery for skull base meningiomas between 1983 and 2008. Among them, 32 (8.2%) showed tumor recurrence. Risk factors for recurrence were analyzed clinically and biologically. Recurrent cases were treated with radiotherapy, surgery, or both. In reoperation cases, pathological and biological changes were analyzed and compared between groups with or without radiotherapy. The recurrence rate was statistically high in cases of partial tumor removal and in patients with tumor in the cavernous sinus, tumors with histological WHO (World Health Organization) grade ≥ II or MIB-1 index > 3. The local control rate of postoperative radiotherapy for recurrent cases was 66.7%. Malignant transformation and MIB-1 index elevation was observed more frequently in patients who underwent reoperation after radiotherapy than in the reoperation-only group. Risk factors for recurrence of skull base meningiomas are as follows: (1) partial tumor removal, (2) tumor in the cavernous sinus, (3) histological WHO grade ≥ II, or (4) MIB-1 index > 3. Postoperative radiotherapy might be effective for tumor recurrence. However, the indications for radiotherapy should be carefully considered because postsurgical radiotherapy may increase biological activity, inducing malignant transformation.

Introduction

Meningiomas account for 15 to 20% of all primary brain tumors. ^{1 2} Almost 20 to 30% of these intracranial meningiomas occur in the skull base. ³ Skull base meningiomas are intimately associated with cranial nerves, vessels, and the brainstem. ⁴ Total surgical removal is the “gold standard” for meningiomas; however, incomplete removal of skull base meningiomas is sometimes practiced in an effort to preserve patients' functionality. In such cases, tumor recurrence or regrowth is a significant problem. Nevertheless, the biological effects of radiotherapy or surgery on skull base meningiomas remain unclear. Herein, we report the clinical and biological effects of radiotherapy and surgery on skull base meningiomas. In addition, we compared biological tumor activity between the first and second operations.

Patients and Methods

The medical records of 392 patients with skull base meningiomas who underwent their initial surgery at Keio University Hospital between 1983 and 2008 were retrospectively reviewed for data collection by our department. Age was defined at which the patient underwent their initial treatment.

Tumor Recurrence

In this report, the term “recurrence” includes both recurrence after gross total removal (GTR) and regrowth after incomplete removal. The recurrence period was defined as the duration between the time of surgery and recurrence or progression which was evaluated on the basis of findings from imaging studies performed twice a year or symptom development, whichever occurred first.

Tumor Locations

Tumors were defined on the basis of their anatomical location. Cranial meningiomas arising from the (1) optic nerve sheath, (2) anterior fossa, (3) sphenoid wing, (4) tuberculum sellae, (5) middle fossa, (6) cavernous sinus, (7) central skull base, (8) petroclival region, (9) cerebellopontine angle, (10) jugular foramen, (11) lower clivus, or (12) foramen magnum were classified as skull base lesions.

Extent of Surgical Removal

The extent of surgical resection was estimated on the basis of operative notes and postoperative imaging findings. The extent of surgical removal was classified as follows: GTR, no residual tumor on operative and postoperative radiological findings; nearly total resection (NTR), ≥ 95% removed; subtotal resection (STR), ≥ 75 and < 95% removed; and partial resection (PR), < 75% removed. Therefore, GTR was in accordance with Simpson's grade 1, 2, NTR and STR were consistent with grade 3, and PR was equivalent to grade 4. The overall removal rate of the tumors was as follows: GTR, 62.5% (245/392); NTR, 9.5% (37/392); STR, 25.0% (98/392); and PR, 3.3% (13/392).

Tumor in the Cavernous Sinus

The presence of tumor invasion into the cavernous sinus was estimated by intraoperative findings and postoperative radiological images.

Histological World Health Organization (WHO) Grade

Histological tumor grade was determined on the basis of information from the pathology reports and classified according to the WHO criteria, 2000. ¹

MIB-1 Index

The MIB-1 index which was included in the medical records starting in 1990 was examined in 332 cases.

Treatment for Tumor Recurrence

Thirty-one patients with recurrent tumor were treated by reoperation, radiotherapy, or both. Radiotherapy modalities included the gamma knife, linear accelerator (LINAC), and conventional radiotherapy; however, since 2000, the gamma knife has been the most frequently used technique.

Fourteen patients underwent only radiotherapy for tumor recurrence, seven patients whose tumors could not be controlled by radiotherapy underwent subsequent reoperation, and 11 patients did not receive any radiotherapy. In the patients who underwent reoperation, both the group that received radiotherapy (RT + OP, n  = 7) and the group that did not (OP, n  = 11), we compared the rate of malignant transformation and MIB-1 index between the initial and second operations. Malignant transformation was defined as WHO grade I to II or IIII and II to III at second surgery.

Statistical Analysis

Patients were classified according to the presence or absence of tumor recurrence and the following seven items were assessed statistically: age, gender, tumor location, extent of surgical removal, presence of the tumor in the cavernous sinus, histological WHO grade, and MIB-1 index. The Chi-square test was used to compare gender, tumor locations, extent of surgical removal, presence of tumor in the cavernous sinus, and histological WHO grade. In patients with recurrence who received postoperative radiotherapy, the Chi-square test was adopted to compare the recurrent rates among the type of radiotherapy. The paired t -test was used to compare the age and MIB-1 index in cases with and without recurrence. In the patients who received reoperation, the paired t -test was used to compare the MIB-1 index between initial and second operations in RT + OP or OP groups. Statistical analyses were performed using Stat View version 5.0 (SAS Institute, Cary, NC, U.S.A.) on a personal computer. Differences were considered significant when p -value was less than 0.05.

Results

The overall rate of recurrence was 8.2% (32/392). The average recurrence period between the initial operation and diagnosis of recurrence was 37.3 ± 5.9 months, and the average follow-up period for nonrecurrent cases was 63.2 ± 3.0 months. The recurrence rates among the various factors are listed in Table 1 .

Table 1. Recurrence rates by case characteristics.

Item	Cases	Recurrence rate (cases)
Total	392	8.5% (32)
Gender
Female	296	8.4% (25)
Male	96	7.3% (7)
Tumor location
1. Optic nerve sheath	15	33.0% (5)
2. Anterior fossa	36	5.6% (2)
3. Sphenoid wing	73	5.5% (4)
4. Tuberculum sellae	33	3.0% (1)
5. Middle fossa	21	9.5% (2)
6. Cavernous sinus	7	14.3% (1)
7. Central skull base	19	5.3% (1)
8. Petroclival region	100	11.0% (11)
9. Cerebellopontine angle	62	4.8% (3)
10. Jugular foramen	5	20.0% (1)
11. Lower clivus	5	0% (0)
12. Foramen magnum	16	6.3% (1)
Extent of surgical removal
GTR	247	5.6% (14)
NTR	35	8.5% (3)
STR	98	10.2% (10)
PR	12	41.7% a (5)
Tumor in the cavernous sinus
Yes	51	19.6% b (10)
No	341	6.5% (22)
Histological WHO grade
I	375	6.7% (25)
II	17	41.2% c (7)

Abbreviations: GTR, gross total removal; NTR, near total removal; PR, partial removal; STR, subtotal removal.

^a Significantly high ( p  < 0.05) compared with GTR, NTR, and STR by Chi-square test.

^b Significantly high ( p  < 0.05) compared with the group having no cavernous sinus tumor by Chi-square test.

^c Significantly high ( p  < 0.05) compared with histological WHO grade II group by Chi-square test.

Age and Gender

The mean age of patients in the recurrence group was 50.8 ± 2.1 years of age (mean ± standard error), which was lower than that of the patients in the nonrecurrent group (53.5 ± 0.7 years); however, there was no statistical intergroup differences. The rate of skull base meningiomas in females (75.5%, 296/392) was about three times of that for males (24.5%, 96/392) but there was no statistical difference in the rate of recurrence between females (8.4%, 25/296) and males (7.3%, 7/96).

Tumor Locations

Recurrence rates by tumor location were as follows: (1) optic nerve sheath, 33.3% (5/15); (2) anterior fossa, 5.6% (2/36); (3) sphenoid wing, 5.5% (4/73); (4) tuberculum sellae, 3.0% (1/33); (5) middle fossa, 9.5% (2/21); (6) cavernous sinus, 14.3% (1/7); (7) central skull bases, 5.3% (1/19); (8) petroclival, 11.0% (11/100); (9) cerebellopontine angle, 4.8% (3/62); (10) jugular foramen, 20.0% (1/5); (11) lower clivus, 0% (0/5); and (12) foramen magnum 6.3% (1/16). There was no statistically significant difference among recurrence rates at the various tumor locations.

Extent of Surgical Removal

The recurrence rate decreased according to the extent of tumor removal: GTR, 5.6% (14/249); NTR, 8.5% (3/35); and STR 10.2% (10/98). The recurrence rate in the PR group was 41.7% (5/12) which was significantly higher than those in the other groups ( p  < 0.05).

Tumor in the Cavernous Sinus

The recurrence rate in patients with tumor in the cavernous sinus was 19.6% (10/51) which was significantly higher than the recurrence rate in patients without tumor in the cavernous sinus (6.5%, 22/341).

Histological WHO Grade

There were no histological WHO grade III cases at initial surgery, in this study. The recurrence rate of the histological WHO grade I group was 6.7% (25/375). The recurrence rate of the histological WHO grade II group was 41.2% (7/17) which was significantly higher than that of the histological WHO grade I group ( p  < 0.05).

MIB-1 Index

Beginning in 1990, we examined the MIB-1 index of skull base meningiomas (recurrent, 31 cases; nonrecurrent, 301 cases). As indicated in Table 2 , the mean MIB-1 index of tumors in the recurrent group was 3.86 ± 0.71 (mean ± standard error) which was higher than that in the nonrecurrent group (2.84 ± 0.14), although not significantly different ( Table 2 ).

Table 2. MIB-1 index in recurrent and nonrecurrent tumors.

	Recurrent tumors ( n  = 31)	Nonrecurrent tumors ( n  = 301)
MIB-1 index (Mean ± SE)	3.86 ± 0.71	2.48 ± 0.14

Abbreviation: SE, standard error.

Local Control Rate after Postoperative Radiotherapy

Twenty-one patients underwent postoperative radiotherapy and the local control rate as total was 66.7% (14/21). The local control rate in the gamma knife group, the LINAC group, and the conventional group were 69.2% (9/13), 66.7% (2/3), and 60.0% (3/5), respectively ( Table 3 ). There was no significant difference among the three groups.

Table 3. Local control rate of postoperative radiotherapy after initial surgery.

	Gamma knife	LINAC	Conventional	Total
Local control rate (cases)	69.2% (9/13)	66.7% (2/3)	60.0% (3/5)	66.7% (14/21)-

Abbreviation: LINAC, linear accelerator.

Biological Changes of the Tumor at Reoperation

Malignant Transformation

As indicated in Table 4 , the rate of malignant transformation in the RT + OP group was 57.1% (4/7), which was higher than that in the OP group (18.2%, 2/11).

Table 4. Biological change of the tumor at reoperation.

	RT + OP ( n  = 7)		OP  ( n  = 11)
Rate of malignant transformation	57.1% (4/7)		18.2% (2/11)
MIB-1 index  (Mean ± SE)	4.22 ± 1.18 (initial)	13.14 ± 3.81 a (second)	4.21 ± 1.55 (initial)	4.48 ± 1.64 (second)

Abbreviations: OP, Reoperation only; RT + OP, Reoperation after radiotherapy; SE, standard error.

^a Significantly increased ( p  < 0.05) compared with that at the time of initial operation (compared by the paired t -test).

Elevation of MIB-1 Index

Changes in MIB-1 index were studied in 17 patients who underwent reoperation. In RT + OP group ( n  = 7), MIB-1 index at the second operation (13.14 ± 3.81) was about three times higher than that at the initial operation (4.22 ± 1.18), and this difference was statistically significant ( p  < 0.05; Fig. 1 ). On the other hand, there was no statistically significant difference between the initial MIB-1 index (4.21 ± 1.55) and that at the second operation (4.48 ± 1.64) in OP group ( n  = 10; Fig. 1 ), suggesting that the tumor was biologically activated by radiotherapy.

Fig. 1.

The graph for the reoperation-after-radiotherapy group shows MIB-1 elevation at the time of the second operation (left). MIB-1 elevation was not observed in the reoperation-only group (right). Abbreviations: OP, Reoperation only; RT + OP, Reoperation after radiotherapy.

Rereccurent Rate

The total rereccurent rata of tumor recurrence was 25.0% (8/32). The rereccurent rate in OP group was 36.3% (4/11) which had no statistical difference in that of RT + OP group (57.1%, 4/7). The patients with rerecurrence underwent additional surgery or radiotherapy. The mortality rate of rerecurrence was 12.5% (4/32: tumor death, three cases; meningitis, one case).

Discussion

The overall recurrence rate of skull base meningiomas in our study was 8.2%. When taking Simpson's grade into consideration, the recurrence rates were 8% for grade 1, 16% for grade 2, 29% for grade 3, and 42% for grade 4. ⁵ Since the publication of Simpson's study, others have confirmed that the extent of surgical resection is an influential factor on local recurrence. ^{6 7 8} Nanda and Vannemreddy reported that the recurrence rate of skull base meningiomas treated with GTR was 26.5%, whereas the recurrence rate of nonskull base meningiomas treated with GTR was 15.0%. ⁴ Jääskeläinen and colleagues reported that the recurrence rate of benign meningiomas was 3% at 5 years and 21% at 25 years. ⁹ Mahmood and associates reported recurrence-free rates of 98% at 5 and 10 years in patients having complete resections of benign meningiomas compared with 54 and 36% in those with incomplete removals at the same intervals. ¹⁰ Mathiesen et al reported a 5-year recurrence rate with GTR (Simpson's grade 1 or 2) of 4% and 25 to 45% with grade 3 or 4. ¹¹ Since we achieved a good removal rate of skull base meningiomas (63.5% GTR), we were able to control recurrence well and our results were similar to those of previous reports. The location of the meningioma tumor influences the extent of surgical resection which in turn influences the rate of reccurence. ⁴ It is important to remove the tumor as radically as possible at the initial operation because the extent of surgical removal at the second surgery seldom exceeds that of the first operation.

De Jesús and colleagues reported a 5-year local control rate of 81% after complete resection compared with 62% after subtotal resection in cavernous sinus meningiomas. ¹² Because entering the cavernous sinus entails a significantly higher risk of complications, radiosurgical treatment should be reserved for remnants with secondary growth and clinical manifestations. ¹³ Any cranial neuropathy with meningiomas involving the cavernous sinus has a good chance of improvement after radiotherapy. ¹⁴ Resection should be limited to the extracavernous extensions and radiotherapy after surgery should be considered only in cases with other risk factors for recurrence.

The 5-year recurrence rate for patients with histological WHO grade I has been reported as 21% while the recurrence rates for patients with histological WHO grade II and III meningiomas have been noted to be 38 and 78%, respectively. ¹⁵ In the present study, the recurrence rate of histological WHO grade II tumors was 41.2% which is compatible with previous reports.

Nevertheless, the histological features of tumors are not always predictive of tumor recurrence, especially in histological WHO grade I meningiomas. It is not possible to determine the proliferative potential of meningiomas based on classical histological classification using hematoxylin-eosin sections. ^{16 17} The use of the monoclonal antibody MIB-1 index to forecast meningioma tumor growth has been reported by several authors. ^{16 18} Previous studies have shown that the proliferation potential of meningiomas, as well as prediction of their natural history, can be based on the MIB-1 index. ^{17 19} The MIB-1 index may, in fact, be labeling actively dividing cells whose daughters undergo a high rate of cell death to balance excessive growth. Matsuno et al and Nakasu et al (1987 and 2001) reported 3.2 and 3% cut-off points, respectively, for a higher recurrence tendency. ^{16 20 21} In our study, there was no statistically significant difference between the MIB-1 index of tumors with and without recurrence, although the mean MIB-1 index of recurrent tumors was 3.86. In patients whose MIB-1 index was over 3, the recurrence-free rate within 5 years was said to be over 40%. ^{16 21}

Although meningiomas have historically been considered to be radioresistant, ^{5 22} analysis of a more recent series suggests that radiotherapy improves local control of meningiomas. ²³ Postoperative radiotherapy is often used to improve local control. Relative to surgery alone, surgery with postoperative irradiation generally achieves an improved recurrence-free interval for both benign and malignant meningiomas. ²² There is a general consensus that the gamma knife, ^{14 24 25} intensity-modulated radiotherapy, ^{26 27 28 29} and proton beam therapy ^{30 31 32} are effective for high histological grade meningiomas or cavernous sinus meningiomas that cannot be removed without morbidity. Stafford et al found that the 5-year local control rate of meningiomas treated with radiotherapy either primarily or subsequent to surgical treatment was 93, 68, and 0% in patients with benign, atypical, and malignant subtypes, respectively. ³³

Malignant transformation of benign meningiomas must also be considered when evaluating treatment outcome. ³⁴ Al-Mefty et al reported that up to 2% of benign meningiomas transform into malignant meninigomas. ³⁵ Others have reported that 28.5% of all recurrent meningiomas were found to be atypical or anaplastic. ⁹ Iwai et al reported that the rate of malignant transformation was 6.5% in the case of skull base meningiomas with radioptherapy. ³⁴ Couldwell et al insisted that growth of skull base meningiomas could be aggressive after failed radiotherapy in some patients and that consistent and extended follow-up evaluations should be performed in all cases of even benign meningiomas after radiotherapy. ³⁶ Ohba et al reported that evaluation of the MIB-1 index might be useful for predicting the malignant transformation of meningiomas. ³⁷ Our data showed that the rate of malignant transformation in RT + OP group (57.1%) was higher than that in the OP group (18.2%). Thus, the indication of radiotherapy after surgery must be considered carefully because radiotherapy may trigger malignant change.

Some authors have reported that the MIB-1 index did not usually change between the initial tumor and recurrence. ^{38 39} On the other hand, elevation of the MIB-1 index in meningiomas treated by radiotherapy has been reported. ⁴⁰ In our study, the elevation in MIB-1 index between the initial and second operation was statistically significant in RT + OP group but not in OP group. Radiation exposure may result in expression of the nuclear protein-DNA (deoxyribonucleic acid) complex recognized by the MIB-1 index in an aberrant phase of the cell cycle. ⁴⁰ Al-Mefty et al reported that an increasing MIB-1 index in recurrent meningiomas might indicate a process of malignant change. ³⁵ Therefore, since radiotherapy may activate the biological nature of benign tumors, indications for its use should be limited to biologically active meningiomas or to tumors showing regrowth.

There are some limitations to the present study. The number of patients with recurrence was limited. In addition, there would be selection bias to choose surgery or radiotherapy after recurrence because selection of treatment was not randomized and based on our experience and knowledge. Since WHO grade and MIB-1 index was not analyzed in the patients controlled only by radiotherapy after recurrence, all the aspect of the effect of radiotherapy could not be analyzed biologically. In addition, it would be possible that RT + OP group was originally aggressive and had high risk for malignant transformation and elevation of MIB-1 index compared with the patients controlled by only surgery or radiotherapy because tumor recurrence could not be controlled by radiotherapy in RT + OP group. It is hoped to increase our cases to accumulate the further evidence.

Conclusion

The high risk factors for recurrence of skull base meningiomas include the following: (1) partial removal of the tumor (PR), (2) existence of residual tumor in the cavernous sinus, (3) tumor histology of WHO grade ≥ II, and (4) MIB-1 index > 3. Postoperative radiotherapy is effective for local tumor control for recurrent tumor; however, because the incidence of malignant transformation due to radiotherapy is not negligible, indications for radiotherapy should be limited to patients exhibiting high-risk factors.